American Government › Section G · Voting and elections › Topic G.2
Voter turnout
Turnout, the share of people who actually vote, decides which voices officials hear most. How it is measured, who turns out, and why others stay home are questions campaigns and reformers argue over every election year.
Why turnout matters
Voter turnout is the share of a population that casts a ballot in an election. Campaigns study it closely because they cannot win with supporters who stay home, so they spend most of their money on people likely to vote. Turnout also shapes policy. If one group votes at much higher rates than another, elected officials have stronger reasons to respond to it. The previous topic showed how people get onto the voter roll; this one asks how many of them, and of everyone eligible, then vote, and why.
Measuring turnout: the denominator
Every turnout rate divides the number who voted by some population, and the choice of that population changes the answer. The voting-age population (VAP) counts everyone 18 or over, including non-citizens. The citizen voting-age population (CVAP), which the Census Bureau uses, counts only adult citizens. The voting-eligible population (VEP), estimated by the University of Florida Election Lab, also removes people barred by felony laws and adds eligible voters living overseas. The smallest base is registered voters. The Census Bureau's report of August 2026 found that 2024 voters were 59.3 percent of the VAP, 65.3 percent of the CVAP and 88.8 percent of registered citizens.
Worked example: one election, three rates
Use the Census figures for 2024: about 154.3 million voters and 173.9 million registered citizens. 1. Turnout of the registered: 154.3 divided by 173.9 is about 0.887, so roughly 89 percent. 2. Size of the CVAP: if 154.3 million is 65.3 percent, the CVAP is 154.3 divided by 0.653, about 236 million. 3. Turnout of the VAP is lower still, 59.3 percent, because the base adds non-citizen adults. 4. Interpret: the registered rate looks high because registering already filters out the least interested. 5. Conclusion: quote a turnout figure only with its base, and compare elections on the same base.
Self-reports and official counts
The Census figures come from the Current Population Survey, which asks people whether they voted. Surveys tend to overstate turnout, because some non-voters say they voted, though the Bureau's practice of counting the 14 percent of the CVAP who gave no answer as non-voters pulls the other way. The Election Lab instead divides official ballot counts by its estimate of the VEP. On that measure, updated in March 2026, turnout was 60.1 percent in 2016, 66.4 percent in 2020 and 64.3 percent in 2024. The 2020 figure was the highest since 1900. Midterm turnout is far lower: 36.7 percent in 2014, 50.0 percent in 2018 and 46.1 percent in 2022.
Who votes
Turnout follows clear patterns. As the topic on engagement in a democracy showed, it rises steeply with education. It also rises with age and income. In 2024, Americans aged 18 to 29 were 19.6 percent of adult citizens but only 15.3 percent of voters, while those 65 and over were 24.6 percent of citizens and 28.1 percent of voters, the Census Bureau found. About half of voters, 48.8 percent, had family incomes of $100,000 or more. Women have voted at higher rates than men in every presidential election since 1980; in 2024 the figures were 66.9 and 63.7 percent. Moving house, which young people do often, lowers turnout because registration must be updated.
The paradox of voting
Why vote at all, when one ballot almost never decides a national election? The economist Anthony Downs raised this puzzle in 1957. William Riker and Peter Ordeshook later wrote the reward from voting as
where p is the tiny chance your vote is decisive, B the benefit if your side wins, C the cost of voting in time and effort, and D the satisfaction of doing your civic duty. Because pB is close to zero, the model says people vote mainly when D outweighs C. That explains why lowering costs, or strengthening a sense of duty and social pressure, can raise turnout.
Why people stay home
Registered citizens who did not vote in 2024 were asked for their main reason. The commonest answers were lack of interest or a belief their vote would not matter, 19.7 percent; being too busy, 17.8 percent; and disliking the candidates or issues, 14.7 percent. Features of the system add costs. Federal elections fall on a working Tuesday, a date Congress set in 1845. Voting is voluntary, whereas Australia, which has required voting in federal elections since 1924, records turnout of about 90 percent of enrolled voters. Americans also face frequent elections, and many contests are not close, which gives campaigns little reason to mobilise.
Making voting easier, and the ID debate
States have cut the cost of voting with early in-person voting and mail ballots. In 2024, about 31 percent of voters voted early in person, 29 percent by mail and 40 percent on Election Day, the Census Bureau reported. Research suggests convenience voting mostly helps people already inclined to vote, so its turnout effects are modest. Rules can also add costs. In Crawford v. Marion County Election Board (2008), the Supreme Court upheld Indiana's photo ID law, 6-3. Supporters of such laws say they deter fraud and build confidence; opponents say they burden eligible voters who lack ID. Studies disagree about how much these laws change turnout.
Worked example: the midterm drop-off
Compare Election Lab VEP turnout of 64.3 percent in 2024 with 46.1 percent in 2022. 1. The gap: 64.3 minus 46.1 is 18.2 percentage points. 2. Cause: midterms have no presidential race, so news coverage and campaign spending are lower and the stakes feel smaller. 3. Who drops out: occasional voters, who are more often young, less educated and less wealthy. 4. Consequence: the midterm electorate is older and more habitual, so candidates stress issues that matter to it. 5. Lesson: a change in who turns out, not just a change of opinion, can swing an election.
Common mistakes to avoid
First, comparing turnout rates calculated on different bases, such as a registered-voter rate against a VEP rate. Second, treating survey self-reports as exact counts: they run high. Third, saying US turnout is always falling; it reached a 120-year high in 2020. Fourth, explaining low turnout by apathy alone, ignoring costs such as registration, work and travel. Fifth, assuming midterm and presidential electorates are the same people.
Key takeaways
Turnout divides voters by a base: VAP, CVAP, VEP or registered voters. In 2024, voters were 59.3 percent of the VAP, 65.3 percent of the CVAP and 88.8 percent of the registered (Census); VEP turnout was 64.3 percent (Election Lab), after 66.4 percent in 2020, the highest since 1900. Midterm turnout is much lower. Turnout rises with education, age and income, and women out-vote men. The calculus of voting, R = pB − C + D, shows why costs and civic duty matter. Convenience voting and ID laws change costs, with contested effects.
Flashcards
Flashcard
The share of a population that casts a ballot
Test your understanding
Question 1 of 5
In 2024 the Census Bureau found that voters were 88.8 percent of one group but only 59.3 percent of another. Which pairing is correct?
Question 2 of 5
Using Election Lab VEP figures, how much lower was turnout in the 2022 midterm than in the 2024 presidential election?
Question 3 of 5
In the calculus of voting, R = pB − C + D, why does the model predict that people vote mainly for reasons of duty or satisfaction?
Question 4 of 5
Why do Census turnout estimates tend to be higher than counts of actual ballots would suggest?
Question 5 of 5
Which statement about who voted in 2024 is supported by the Census figures?
